Understanding Scientific Assessments in Adversarial Polities through Quebec Sh

154 Pages·2016·1.87 MB·English

Université de Montréal
An Enforced Cooperation: Understanding Scientific Assessments in
Adversarial Polities through Quebec Shale Gas Policymaking, 2010-2014
Par Alexandre Harvey
Département de Science politique, Faculté des Arts et des Sciences
Mémoire présenté en vue de l'obtention du grade de Maîtrise ès sciences
en Science politique
Avril 2016
© Alexandre Harvey, 2016
Résumé
Les biotechnologies, le réchauffement climatique, les ressources naturelles et la gestion des écosystèmes
sont tous représentatifs de la “nouvelle politique de la nature” (Hajer 2003), un terme englobant les enjeux
marqués par une grande incertitude scientifique et un encadrement réglementaire inadapté aux nouvelles
réalités, suscitant de fait un conflit politique hors du commun. Dans l'espoir de diminuer ces tensions et de
générer un savoir consensuel, de nombreux gouvernements se tournent vers des institutions scientifiques ad
hoc pour documenter l'élaboration des politiques et répondre aux préoccupations des partie-prenantes. Mais
ces évaluations scientifiques permettent-elles réellement de créer une compréhension commune partagée par
ces acteurs politiques polarisés? Alors que l'on pourrait croire que celles-ci génèrent un climat
d'apprentissage collectif rassembleur, un environnement politique conflictuel rend l'apprentissage entre
opposant extrêmement improbable. Ainsi, cette recherche documente le potentiel conciliateur des évaluation
scientifique en utilisant le cas des gaz de schiste québécois (2010-2014). Ce faisant, elle mobilise la
littérature sur les dimensions politiques du savoir et de la science afin de conceptualiser le rôle des
évaluations scientifiques au sein d'une théorie de la médiation scientifique (scientific brokerage). Une analyse
de réseau (SNA) des 5751 références contenues dans les documents déposés par 268 organisations participant
aux consultations publiques de 2010 et 2014 constitue le corps de la démonstration empirique. Précisément, il
y est démontré comment un médiateur scientifique peut rediriger le flux d'information afin de contrer
l'incompatibilité entre apprentissage collectif et conflit politique. L'argument mobilise les mécanismes
cognitifs traditionnellement présents dans la théorie des médiateurs de politique (policy broker), mais
introduit aussi les jeux de pouvoir fondamentaux à la circulation de la connaissance entre acteurs politiques.
Mots clés : médiation, évaluation scientifique, apprentissage collectif, élaboration des politiques publiques,
conflit politique, analyse de réseau, sous-système, Exponential Random Graph Model
Summary
Biotechnology, climate change, natural resources, and ecosystem management are all representative of
the “new politics of nature” (Hajer 2003), a term encompassing policy issues with high scientific
uncertainties, unadapted regulatory regimes, and acute political conflict. In the hope of diminishing these
tensions and generating a consensual understanding, several governments mandated ad hoc scientific
institutions to document policymaking and answer stakeholder’s concerns. But do those scientific
assessments really help to generate a shared understanding between otherwise polarized policy actors? While
it would be possible that these create inclusive collective learning dynamics, policy learning has been shown
as being extremely unlikely among competing policy actors. Accordingly, this research documents the
conciliatory power of scientific assessments using the Quebec shale gas policymaking case (2010–2014). In
doing so, it mobilizes the literature stressing the political nature of science to conceptualize scientific
assessment in light of a scientific brokerage theory. Empirically, the research uses Social Network Analysis to
unravel the collective learning dynamics found in two information networks built from the 5751 references
found in the advocacy and technical documents published by 268 organizations during two public
consultations. Precisely, findings demonstrate that scientific brokerage can redirect information flows to
counteract the divide between collective learning and political conflict. The argument mobilizes cognitive
mechanisms traditionally found in policy brokerage theory, but also introduces often forgotten power
interplays prominent in policy-related knowledge diffusion.
Keywords: brokerage, scientific assessment, collective learning, policymaking, policy, political conflict,
social network analysis, subsystem, Exponential Random Graph Model
Table of Contents
List of Figures..........................................................................................................................................v
List of Tables...........................................................................................................................................vi
List of Abbreviations.............................................................................................................................vii
Introduction.............................................................................................................................................1
I - An Important Inquiry...................................................................................................................................2
II - The Study in Theoretical Perspectives........................................................................................................4
III - Methodological Strategy...........................................................................................................................6
IV - Structure of the Thesis...............................................................................................................................8
Chapter 1: Theoretical Framework.......................................................................................................9
I - Understanding Policy-oriented Knowledge..................................................................................................9
Political Knowledge.....................................................................................................................................................10
A Review of Collective Learning..................................................................................................................................12
An Integrated Picture of Collective Learning..............................................................................................................21
II – A Network-based Approach to Collective Learning.................................................................................22
Coalitions and belief systems.......................................................................................................................................23
Other Actors..................................................................................................................................................................24
Expected Dynamics of Adversarial Subsystems...........................................................................................................26
III – Brokering a Crisis Recovery...................................................................................................................29
Delineating the Broker: A Definition............................................................................................................................29
How Brokerage Improves Policy Networks: The Functions........................................................................................32
How Brokers Reach Structural Holes: Acquiring Influence........................................................................................39
A Hypothesis.................................................................................................................................................................46
IV - Summary.................................................................................................................................................46
Chapter 2: Case and Methodology......................................................................................................48
I - The politics of shale gas.............................................................................................................................49
Scientific Controversies................................................................................................................................................49
From Science to Quebec Politics..................................................................................................................................53
From Events to Research Design..................................................................................................................................59
i
II – Data..........................................................................................................................................................60
Relational Data: From References to Collaborative Dynamics...................................................................................61
III - Analytical Strategy: A Dual Investigation................................................................................................64
A Macro-Order Analysis...............................................................................................................................................64
A Micro Order of Analysis............................................................................................................................................69
IV - Conclusion..............................................................................................................................................73
Chapter 3: Results and Discussion......................................................................................................75
I - Improved Collective Learning....................................................................................................................75
II – Influential Brokerage...............................................................................................................................78
An Explanation: The Collaborative Core Thesis..........................................................................................................84
Inferential Modelling....................................................................................................................................................88
III - Discussion: Strengthening Brokerage Theory..........................................................................................92
Explaining the Collaborative Core...............................................................................................................................93
A Better Definition of Policy Brokers...........................................................................................................................95
The State of Information in Adversarial Policymaking................................................................................................97
IV - Conclusion..............................................................................................................................................99
Concluding Remarks...........................................................................................................................100
I - Addressing Limits and Their Consequences.............................................................................................101
Capturing Collective Learning...................................................................................................................................101
Network Construction.................................................................................................................................................103
ERGM Degeneracy.....................................................................................................................................................104
External Validity.........................................................................................................................................................105
II - Few Words on a Brokerage Research Agenda.........................................................................................106
Appendix A: SNA Statistics Details...................................................................................................108
Reciprocity...................................................................................................................................................108
Burt's constraint............................................................................................................................................108
Hubs and Authorities Algorithm...................................................................................................................109
Similarity Matrices.......................................................................................................................................109
Louvain algorithm........................................................................................................................................110
Closeness......................................................................................................................................................111
Local transitivity...........................................................................................................................................111
ii
Exponential Random Graph Model...............................................................................................................111
Appendix B: Degeneracy and Goodness of-Fit.................................................................................112
MCMC Behaviour........................................................................................................................................112
Goodness-of-fit.............................................................................................................................................119
Appendix C: R codes...........................................................................................................................123
Bibliography........................................................................................................................................130
iii
List of Figures
Figure 1.1: A Comprehensive Framework of Collective Learning.........................................................14
Figure 1.2: Structural Holes of Policy Network......................................................................................34
Figure 1.3: Models of Trust-building......................................................................................................37
Figure 1.4: Structural Advantages of Brokers for Information Circulation............................................38
Figure 1.5: The Framework Applied to Information Networks and Scientific Brokerage......................44
Figure 2.1: Horizontal Fracturing and Shale Gas Extraction..................................................................51
Figure 2.2: Montly Coverage of Shale Gas in the Province of Quebec, 2010-14...................................54
Figure 2.3: Distribution of Policy Preferences........................................................................................56
Figure 2.4: Brokerage Types...................................................................................................................70
Figure 2.6: The Relation Between Authority Score, Hub score, and Brokerage....................................71
Figure 3.1: Distribution of Normalized Authority in 2011 and 2014......................................................77
Figure 3.2: Brokerage Count in 2011 and 2014......................................................................................79
Figure 3.3: Authority and Hub Scores.....................................................................................................82
Figure 3.4: Distribution of Organizational Affiliation by Information Community...............................83
Figure 3.5: Cumulative Authority of Information Sources by Community............................................87
Figure 4.1: MCMC Estimation Behaviour [1/3]...................................................................................113
Figure 4.2: MCMC Estimation Behaviour [2/3]...................................................................................114
Figure 4.3: MCMC Estimation Behaviour [3/3]...................................................................................115
Figure 5.1: In-degree Goodness-of-fit...................................................................................................119
Figure 5.2: Out-degree Goodness-of-fit................................................................................................120
Figure 5.3: Shared-partners Goodness-of-fit.........................................................................................121
Figure 5.4: Minimum Geodesic Distance Goodness-of-Fit..................................................................122
iv
List of Tables
Table I: Summary of the Macro-order Instruments.................................................................................69
Table II: Summary of the Micro-order Instruments................................................................................73
Table III: Summary of Collective Learning Dynamics...........................................................................76
Table IV: Collective Learning Dynamics in Assessment and Advocacy Subnetworks..........................85
Table V: ERGM Results..........................................................................................................................89
Table VI: Typical Cases Facilitating ERGM Interpretation....................................................................91
Table VII: General Estimation Information..........................................................................................116
Table VIII: Descriptive Statistics for Each Variable.............................................................................116
Table IX: Cross-correlation...................................................................................................................117
Table X: Auto-correlation......................................................................................................................118
v
List of Abbreviations
ACF Advocacy Coalition Framework
BAPE Bureau d'audience publique sur l'environnment
EES Évaluation environnementale stratégique
ERGM Exponential Random Graph Model
IPCC Intergovernmental Panel on Climate Change
SNA Social Network Analysis
vi
Introduction
Climate change mitigation, biotechnology, and ecosystem management are all
representative of the “new politics of nature” (Hajer 2003). Characterized by an increasing
complexity, they regularly exhibit an important degree of technicality, ambiguity, and
uncertainty. Their scope goes well beyond the human one and possesses a multidimensional
nature. Such peculiarities often lead to a broad range of individuals to involve in the
policymaking process, each of them having their own—and most of the time incompatible—
set of political preferences, worldviews, and interests. Those actors respective expertise are
diverse and range from a user-gathered local knowledge to an impressive amount of scientific
sophistication in a delimited field of inquiry. As a consequence, post-materialist issues such as
those noted above are frequently entangled with acute political contention resulting from
competing interests.
From a governmental point of view, the new politics of nature poses important
challenges to good management. The significant amount of complexity induces legitimacy
and analytical capacity concerns, which in turn push governments toward a broader
articulation of governance to achieve efficient problem solving. Such articulation of
governance, however, falls outside traditional policymaking institutions. Indeed, issues of the
new politics of nature transcend existing structures of governance and occur through what
Hajer called the “institutional void”:
“established institutional arrangements often lack the powers to deliver the required or
requested policy results on their own [and, accordingly, policymaking takes] part in
transnational, polycentric networks of governance in which power is dispersed” (2003,
175).
In the absence of agreed rules, the polity becomes mostly informal and emerges as a
consequence of deliberative policymaking instead of constraining it. Understood
simultaneously, the interaction between (1) the need for a decentralized form of governance
(2) high potential for political conflict, and (3) deliberative and poorly institutionalized
1
political arenas poses crucial political challenges to contemporary societies.
To grasp those challenges, numerous societies rely on broad, open, and deliberative
scientific assessments to document the issue at hand and give political actors a common
understanding of a given phenomenon. However, insights from the new politics of nature
suggest the probabilities of success are rather thin. As explained with greater details in the
theoretical part of this research, political conflict, deliberation, and genuine collective learning
make poor associates. Strong policy-related beliefs induce political confrontations, which in
turn activate cognitive defence mechanisms further limiting the potential for constructive
dialogue. Hence, understanding the processes by which societal debates become polarized and
means of mitigating such outcome rises as a pressing and necessary research agenda.
Assessing how information is created, how it circulates across policy actors, and why a
particular argument is rejected or welcomed into an individual’s cognitive system constitutes
a significant contribution to the new politics of nature. Specifically, this research contributes
to this research agenda by investigating the capacity of scientific assessments to diminish
political conflict:
Can scientific assessments create a shared understanding between
political actors operating in adversarial policymaking?
I - An Important Inquiry
While one should acknowledge that the authority of science and expertise in
policymaking has been severely criticized in the literature to the point where some
practitioners developed their own skepticism about their validity (Jasanoff and Wynne 1998),
reasons are strong to believe that scientific knowledge remains central to policymaking and
still holds a firm intellectual leverage and legitimacy-building potential. From an empirical
perspective, science is frequently invoked on pragmatic grounds as the predominant technique
to reduce uncertainty. Shale gas, which constitutes the empirical foundation of this research,
remarkably illustrates such phenomenon. Heavily concerned by the environmental, social, and
2